search

sigmf / SigMF

The Signal Metadata Format Specification
Creative Commons Attribution Share Alike 4.0 International
348 stars 75 forks source link

Geolocation in captures/global/annotations #278

Open 777arc opened 1 year ago

777arc commented 1 year ago

We had a request that geolocation also be allowed in captures for moving receivers that get a periodic location update. This brought up the argument that it might as well only be in captures, just like frequency, the only downside would be backwards combability issues with tooling already pulling it from global. Tooling designed to only have 1 coordinate for a Recording (e.g. a maps interface on a tool) can just pull it from captures[0].

On a related note, it would be nice to have a way for an annotation to contain a geolocation of the emission associated with that annotation (the transmitters location, not receiver). This could be results from TDOA or it could be "label data" known beforehand, etc. This one seems a lot simpler but should it be part of an extension or the core spec?

jacobagilbert commented 1 year ago

Captures scope geolocation is probably reasonable, though we cannot remove the global scope field for a while.

Annotations (where geo info used to be) should probably go into the spatial extension.

777arc commented 1 year ago

"for a while" how does that even work, e.g. would we mark it as deprecated or something?

Yep I'm on board with using the spatial extension for annotations

jacobagilbert commented 1 year ago

Sorry, that was pretty vague. I think if global was undesirable we could formally deprecate in 2.0 and then remove in 3.0? Adding this to captures seems reasonable though.

777arc commented 1 year ago

Works for me, I'll make a PR

jacobagilbert commented 1 year ago

Thanks, can you do these separately? the spatial one is easy but the other may take some time